INTRODUCTION TO THE CORPUS

First of all, a description of my corpus to start with my portfolio for computational musicology. I chose for one of the biggest Dutch Hip Hop artists of the past decade, Sevn Alias, as the main focus of my corpus, with the related question: ‘’Which Sevn Alias albums were the most commercially successful and does this commercial success correspond in any way to the label where the artist was at the time signed at?’’. I chose this corpus of all his Spotify available solo-albums because Sevn Alias was one of the first Dutch hip hop artists who started popping up at the beginning of the rise of the influence of hip hop in the Dutch music scene, and he is considered to be one of the biggest and most influential (based on his music and the cultural influence around it). The main interesting part of it is, in my opinion, the fact that he released over 10 albums in the past decade, all with major success and most of them distributed under different labels in each of the release periods. The natural groups and points of comparison in the corpus are mainly focussed on the artist Sevn Alias his solo albums, the music labels who distributed the albums and the amounts of streams of each track from every album. I think these groups are most relatable on the point that it is the same artist on every track and differ the most on the producers behind the tracks and possible featurings on the tracks. Although, these can also be the main points that I’m still a bit uncertain about, because they also have an influence on the outcome of each track and therefore, the possible commercial success. I created my own Spotify-playlist with all available tracks from every solo-album of Sevn Alias available on solely Spotify. The following link leads to the public playlist I created: https://open.spotify.com/playlist/36UKFGHMewsc9n8KIspNrB?si=bdfe0cdf208347fb

In relation to the structure of the corpus, there are a couple of sidenotes i have to adress. First of all, his very first solo album ‘TwentyFourSevn’ is not available on Spotify, so this won’t be applicable in the research. Secondly, the track called ‘Intro’ from the album ‘Trap Grammy’ is also not applicable, since it gave an error rendering some of the graphs related to the corpus since there is one more track with exactly the same track name.

Last of all, a quick overview of the albums and the labels who distributed them, in order of release date (2016 - 2022).

Twenty Four Sevn 2 - Rotterdam Airlines

Twenty Four Sevn 3 - Rotterdam Airlines, Cloud 9

Trap Grammy - Rotterdam Airlines, Sony Music

Picasso - Rotterdam Airlines

OAKINN - 1OAK Entertainment

Recasso - 1OAK Entertainment

SIRIUS - 1OAK Entertainment

Twenty Four Sevn 4 - 1OAK Entertainment

Bogey 4 the Win - 1OAK Entertainment

Six Summers - 1OAK Entertainment

The Last Flowbender - 1OAK Entertainment

‘At the least half of the ’1OAK Entertainment’ releases in joint venture with Top Notch Music, Universal and the most recent half fully independent as his own label’

FIRST IMPRESSIONS

For my first Spotify API impressions of the corpus, I chose to create several ggplots, since I think that three track features are the support for my topic. In this case the danceability, energy and track popularity API features. I think the track features danceability and energy could play an important factor in terms of commercial success, because I think these factors are important for things like the extent to which it is played in public spheres. A higher danceability and energy level could perhaps support the definition of commercial success per album. The track popularity (number between 0 and 100) represents the overall popularity of the song by album, with a number closer to 100 being Sevn Alias most popular song.

By first impression, we can see that the album ‘Trap Grammy’ is one of the least popular albums and the most recent album ‘The Last Flowbender’ is one of the most popular. If we relate this to both albums general danceability and energy level, we see some interesting aspects. The overal energy and danceability levels of both albums seems to be at the same level, but the album ‘Trap Grammy’ has a wider spread of tracks with some higher energy and danceabilty. This seems to be relateable that the listeners of Sevn Alias’ music relate the track popularity more to tracks with less danceability. Maybe they relate more to a specific style of Rap in the music, than music with high energy and they can dance to.

if we relate the track popularity gram to the albums of the corpus based on streams, we see that it has little relation to the amount of streams an album has to this point in time. For example the album ‘Picasso’ has songs with over 56 million streams and a minimum of 2.5, but the latest album ‘The Last Flowbender’ a track over 1.4 and a minimun of 198.000. Eventhough it shows in the gram that the median track populariy is a lot higher in the latest solo album.

I also find it interesting that the energy and danceability levels are somewhat different in comparison, especially in the releases between 2018 and 2019, where a higher degree of danceability corelates to a lower energy. It is also interesting that the general track popularity of the songs has a relatively general point per album, despite huge hits, ‘outliers’, in for example the year 2018. Around 2021, the album has a broader spread in terms of general popularity. I mainly found these aspects interesting.

VISUALISATION TRACK POPULARITY

VISUALISATION ENERGY

VISUALISATION DANCEABILITY

NOTICEABLE OUTLIERS

Layhoun, enige in E Major, extreem lage track popularity van 15. Centen, enige in B major, extreem lage track popularity van 14. beide van het album TwentyFourSevn 3.

VISUALISATION CHROMAGRAM LAYHOUN

VISUALISATION CHROMAGRAM CENTEN

TIMBRE OUTLIER

Als we een vergelijking maken in instrumentalness en klank kleur (timbre) zien we dat Ik Trap van het album OAKINN een outlier is met de meeste. de tweede meeste is Klim van het album SIRIUS, wat op zich ook een grote outlier vormt. beiden ongeveer evenveel danceability van gemiddeld 0.8, maar verschil in energy - ik trap met 0.7 en klim met 0.4. heeft bijdrage aan de klankkeur daarom timbre cepstogram en self-similarity op beide twee.

VISUALISATION INSTRUMENTALNESS

VISUALISATION IK TRAP CEPSTOGRAM

VISUALISATION IK TRAP SELF-SIMILARITY MATRIX

VISUALISATION KLIM CEPSTOGRAM

VISUALISATION KLIM SELF-SIMILARITY MATRIX

TRACK-LEVEL SUMMARIES

Ten eerste chordograms van de nummers Mileage van album Six Summers met hoogste valence van 0.962 en Stappen van album TwentyFourSevn 2 met laagste valence van 0.123. opmerkelijke verschillen vergelijken . ten slot visualisaties tussen meest en minst populaire albums op basis van gemiddelde track popularity de meest recente albums, … en … op basis van tempo, duration, loudness en visualisatie van timbre coefficients.

VISUALISATION MILEAGE CHORDOGRAM

VISUALISATION STAPPEN CHORDOGRAM

VISUALISATION ALBUMS TEMPO, LOUDNESS, DURATION

VISUALISATION ALBUMS TIMBRE COEFFICIENTS

Tempograms

For this weeks homework I made two tempograms based on two outliers from the corpus. I compared the tracks with the highest and lowest bpm and made both tempograms. The track from the corpus with the lowest bpm is Mileage with a 61.778 bpm, the highest is Middle with 199.928 bpm. PS I have some major technical difficulties with the loading and knitting of my tempograms.

Mileage

Middle

Corpus

For this weeks homework I made a dendogram based on all the individual tracks from the corpus to see how the tracks from all the different albums relate to each other.

Dendogram

CONCLUSION OF MY RESEARCH